研究目的
To propose a new method for fault detection in trackers for PV systems based on pattern recognition analysis, avoiding the use of sensors and a wide range of data.
研究成果
The proposed pattern recognition method is able to identify a fault in the tracker associated to a PV panel. The results show that the method allows the discrimination between the several PV panels that appear in a picture and gives the inclination of all of them. Through the comparison of the several inclination angles of the several panels, the identification of the panels in which there is a fault can be made.
研究不足
The classification accuracy is influenced by image spatial resolution and by image quantization. Increasing the distance between the PV modules and the camera, the finer spatial resolution is required to classify successfully each PV cell. For larger distances between the camera and the PV modules, using a lower spatial resolution images and with a contrast reduction make it difficult to detect separately each PV cell of the PV module.
1：Experimental Design and Method Selection:
The method is based on pattern recognition analysis of photographs of PV modules. The orientation of the PV modules is determined using the centroid of the PV cells after applying an image pre-processing stage. The angle is calculated using statistical moments or by the slope of the line joining two centroids of the PV cells located at the vertices of the PV module.
2：Sample Selection and Data Sources:
48 RGB images of three PV modules with different orientations in both X and Y axis were used. These images were obtained from three panels with manually orientation possibilities.
3：List of Experimental Equipment and Materials:
PV modules, digital camera for image acquisition, and Matlab program for simulation results.
4：Experimental Procedures and Operational Workflow:
The method involves pre-processing the acquired image, main image segmentation, image post-processing, feature extraction, and feature-based classification. The orientation of each panel is determined by the average value of the orientations of the cells that belong to it.
5：Data Analysis Methods:
The orientation of the PV modules is determined through the use of statistical moments or by the slope of the line joining two centroids of the PV cells that are located at the vertices of the PV module. The identification of the PV module under fault is made through the comparison between the orientations of the several panels.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
